A Review of Johnjoe Mcfadden's Book ``Quantum Evolution''

BOOK REVIEW * Quantum Evolution by Johnjoe McFadden reviewed by Matthew J. Donald The Cavendish Laboratory, JJ Thomson Avenue, Cambridge CB3 0HE, Great Britain. e-mail: [email protected] web site: http://people.bss.phy.cam.ac.uk/~mjd1014 J. McFadden, Quantum Evolution: Life in the Multiverse (HarperCollins, 2000), 338 pp, ISBN 0-00-255948-X, 0-00-655128-9. In \Quantum Evolution", Johnjoe McFadden makes far-reaching claims for the importance of quantum physics in the solution of problems in biological science. In this review, I shall discuss the relevance of unitary wavefunction dynamics to biological systems, analyse the inverse quantum Zeno effect, and argue that McFadden's use of quantum theory is deeply flawed. In the first half of his book, McFadden both discusses the biological problems he is interested in solving and gives an introduction to quantum theory. This part of the book is excellent popular science. It is well-written, competent, and fun. As far as the biology is concerned, McFadden, who is a molecular microbiologist, has very specific, and often controversial, opinions. Nevertheless, he does refer to a wide range of alternative points of view. He certainly managed to convince me that I had swallowed too easily the prevailing dogma (as found, for example, in chapter 1 of Albert et al. 1989) about the earliest (\prebiotic") stage in biological evolution during which the first self-replicating molecules appeared. McFadden argues that this stage seems to have happened quite fast in terms of the age of the Earth (perhaps within 100 million years) but that none of the proposed mechanisms, of which he discusses several, give anything like a complete and plausible picture of how to go from the early Earth's chemistry to the first cell. Rescuing the prevailing dogma would require a suitable sequence of laboratory experiments, or, at the very least, plausible computer simulations. Unfortunately, while 100 million years is quite short in the age of the Earth, it is rather long in the laboratory. When McFadden moves on to describe quantum mechanics and its interpreta- tions, I feel that his touch becomes rather less sure. However it is hardly surprising that I can tell that he is out of his primary field and into mine, and the outlines of his presentation still strike me as reasonable. Unfortunately it is the details which matter when he attempts to apply quantum theory. He is not sufficiently explicit about what * quant-ph/0101019, January 2001. In quant-ph/0110083, Johnjoe McFadden and Jim Al-Khalili reply to this review. My response to their reply is at http://people.bss.phy.cam.ac.uk/~mjd1014/qevrevr.html 1 is going on at the level of the quantum state. In my opinion, this eventually leads him hopelessly astray. As well as the appearance of self-replication, McFadden is interested in the possi- bility of \adaptive" or \directed" mutation. This is the claim, reviewed by Lenski and Mittler (1993), that there is some evidence that, in some bacteria, some mutations will tend to appear more frequently in circumstances when they are advantageous than when they are biologically neutral. McFadden also sketches a version of the idea that free will is a quantum phenomenon. In all three cases, the technical core of McFadden's proposals involves the quantum dynamics of molecular systems. The analysis of quantum dynamics is not entirely straightforward. According to the traditional account, the wavefunctions of quantum systems change in two quite different ways. Sometimes there is abrupt, indeterministic, change { \wavefunction collapse" { described by the \projection postulate" in which, with a probability given by a squared transition amplitude, the wavefunction is replaced by some eigenfunction of a \measurement operator". At all other times, the wavefunction changes according to a Schrödingerequation, so that the dynamics is defined by a unitary group of the form U(t) = exp(−itH) with H a self-adjoint Hamiltonian operator. More modern accounts invoke decoherence theory (Giulini et al. 1996) in order to justify the assumption that there is always some sufficiently large scale on which the dynamics is unitary. In particular, decoherence theory encourages us to treat \measurements" as physical processes governed by appropriate global Schrödinger equations and to interpret the abrupt changes as merely the way in which changes in one part of the global quantum system would appear to another part. Although decoherence theory does leave open crucial conceptual problems in the interpretation of quantum theory, it suggests that those problems are not primarily matters of dynamics. The Schrödingerequations which apply in biological situations are precisely those of conventional quantum chemistry and are defined ultimately by the electromagnetic interactions between electrons and nuclei in a fixed background classical gravitational field. Certainly if, as he claims, McFadden's ideas are com- patible with the many-worlds interpretation of quantum theory, then we should also be able to analyse his work in terms of global unitary non-relativistic electromagnetic quantum dynamics. In general, however, subsystems of a system with unitary dynamics will not themselves have unitary (reversible) dynamics unless they are ef- fectively isolated from their surroundings. Moreover, subsystems of a system in a pure (wavefunction) state will not necessarily themselves occupy pure states. Indeed, any account of local states entirely in terms of product wavefunctions, even if it is possible at a single instant, will disregard important issues of thermal physics which are particularly relevant to the apparent dynamics of biological systems. Consider then McFadden's proposals about the process by which self-replicating proteins appear. (McFadden does mention that according to the prevailing dogma self-replication in RNA arose before self-replication in proteins, but he argues that RNA is completely implausible as a prebiotic chemical.) Assume that the early Earth 2 produces a sea of amino acids which can link into peptides a few of which are self- replicating. According to McFadden, the chance of a random peptide being self- replicating is far too small for this process to be a plausible source of life in the framework of classical biochemistry. He claims that, instead, we must invoke what he calls the \inverse quantum Zeno effect”. (Thorough introductions to quantum Zeno effects are given in chapter 8 of Namiki, Pascazio, and Nakazato 1997 and in section 3.3.1 of Giulini et al. 1996.) The idea of the inverse quantum Zeno effect goes back to von Neumann (1932, section V.2). For a mathematician, it is some version of the theorem proved in the appendix which shows that, up to analytical niceties, if the projection postulate is true, and if we can choose our measurements at will, then, by using the right dense sequences of measurements, we can turn some given pure state jΦ><Φj into any other pure state jΨ><Ψj regardless of the physical Hamiltonian. In order to prove this, we first need to use some form of the mathematically- simple existence lemma (A.1) which states that there exists a bounded self-adjoint Hamiltonian K such that exp(−iK)Φ = Ψ. However, the problem with such an existence lemma is that K is totally artificial. It is a purely mathematical adjunct required to reduce to the theorem for the standard Zeno effect in which we have Ψ = Φ. The crucial fact on which the inverse Zeno effect depends, is that, for any pair of unitary propagators U(t) and V (t) defined by suitable Hamiltonians, j<ΦjU(1=N)V (−1=N)jΦ>j2 = 1 − O(1=N 2): (1) In theorem A.2, this fact appears as (A.3) with V (t) = e−itK . (1) is an extension of the principle that unitary change in quantum theory begins slowly. Indeed, using N ×(1=N 2) = 1=N, the idea is that if we can interrupt a change sufficiently often then we can gradually alter it in any given direction. To make use of this, however, we actually do have to have large numbers of interruptions applied within a bounded time. It should also be noted that the interruptions, or projections, constructed in the proof are strongly dependent, through lemma A.1, on the initial and final wavefunctions. Decoherence theory tells us that the projection postulate can be viewed as a phenomenological aspect of a unitary dynamics at a large scale. For a preliminary version of how this works at a mathematical level, we can use another existence lemma similar to lemma A.1. What lemma A.4 means in words is that the projection postulate can always be modelled as the restriction of a unitary dynamics on a large space to density matrix (or mixed state) dynamics on a pair of smaller spaces. Once again, however, the Hamiltonian L in lemma A.4 is totally artificial. For N each of the N projections Pn in theorem A.2 we need to have access to an auxiliary space. We also require that the time s for which the corresponding map exp(−isL) should apply should satisfy s ∼ 1=N and that the dynamics should be switched off at the end of that brief period. The famous experiments of Itano et al. (1990) involved not only a series of brief interactions with auxiliary spaces of photon states, but also a system in which there was an essentially thermodynamical reason for rapid relaxation 3 onto the projected state. Thus there were very specific constraints on the form of the physical dynamics and on the projections. As a physical example of the inverse quantum Zeno effect, McFadden considers three polarizing lenses. Adjust lenses 1 and 2 to block light entirely. Then insert lens 3 between lenses 1 and 2.

Load more